The relation between the tendency to write an online customer review and the discrepancy of a positive or negative experience in a cultural perspective.

(1)

1

The relation between the tendency to write an

online customer review and the discrepancy of

a positive or negative experience in a cultural

perspective.

Sieta Paulusma

Name Sieta Paulusma

Student number S2193841

Date of birth 27-08-1992

E-mail address s.c.paulusma@student.rug.nl

University University of Groningen

Faculty Economics and Business

Study Program Msc Marketing Management & MSc International

Business Management Supervisor (Msc Marketing) dr. J.A. (Liane) Voerman Second supervisor (MSC IBM) dr. R.W. (Rudi) de Vries

(2)

Abstract

In this research we will get a better understanding about the motivations behind a consumer’s decision to engage in EWOM, and especially the tendency to write an OCR. The tendency to write a review is a motivational driver to write an OCR for different reasons, which will lead to different types of OCRs . These motives to engage in EWOM are also effected by emotions of people (Berger and Milkman, 2012). Many researchers did already research in the field of motives to write OCRs which entails for instance; achievement expressing, self confirmation and achieve a therapeutic feeling (Cheung et al. 2010) or product-involvement,

self-involvement (self enhancement), other-self-involvement (concern for others), and

message-involvement (Dichter, 1966). Based on several previous studies, we examined in this research three variables which will be relevant to people’s tendency to write an OCR: concern for others, self-enhancement and feedback for the company. In this research we investigate which tendency people have for writing an OCR, and if this tendency depends on the valence

(positive/negative) and the discrepancy (low/high) of the customer experience. And we will find some differences between the cultural dimensions individualism/collectivism

(IND/COLL) and masculine/feminine (MAS/FEM). This by using German, Chinese and Dutch participants for the research. The results show that there is a significant positive effect of the negative valence of a customer experience on the tendencies to write an OCR.

Discrepancy and the interaction between valence and discrepancy showed no significant effects in the analysis. Also IND/COLL has no significant effects with the three tendencies to write OCRs, this in contract with MAS/FEM what has a positive significant effect with CFO, FC and SE. For marketing managers it will be interesting which type of discrepancy lead to which tendency to write an OCR. So, the get more knowledge about the motivation behind the customer reviews. This knowledge is important because the posted OCRs are often basis for other customers to base their purchase decisions on (Zhu & Zhang, 2010), and this has an effect on the future sales of the company (Forman et al., 2008).

(3)

People often want to share opinions, news, experiences and information with others. They complain about the delay of the train, chitchat about their last bought products of discusses good or bad restaurants. This interpersonal communication is defined as word-of-mouth communication; “informal communications directed at other customers about the ownership, usage, or characteristics or particular products or services of their sellers” (Westbrook, 1987). Word-of-mouth (WOM) includes face-to-face discussions as well as so called “word-of-mouse”, or online reviews (Berger, 2014). Word-Of-Mouth has been recognized as one of the most influential transmission of information, especially for experience goods (Godes & Mayzlin, 2004; Keller, 2007). WOM, and especially online forum information, is perceived to be more credible, trustworthy and relevant than marketer-generated online content (Bickart & Schindler, 2001). With the help of the internet, everyone can share their thoughts, opinions and experiences with other internet users and influence and inspire each other through online, electronic word-of-mouth (EWOM). EWOM is defined as the negative or positive statements about products, services and/or a company, created by co-consumers (potential, actual, or former) who want to inform and advice their peers (Hennig-Thurau et al., 2004; Park & Park, 2008). EWOM communication can take place in many forms, for instance; discussion forums, Web-based opinion platforms, blogs/flogs, news groups, customer reviews, etc. In this study we focus on online customer reviews (OCRs), as these reviews are one of the most powerful channels to generate online word-of-mouth (Duan, Gu & Whinston, 2008).

1.1 Online customer reviews

OCRs were found to successfully influence consumers’ information search, product attitude and purchase intention (Cheung & Thadani, 2010; Doh & Hwang, 2009). It may serve as decision-making aids when consumers are faced with abundant product information and alternative choices (Xie, Miao, Kuo, & Lee, 2011). So, we can conclude, that EWOM, and especially OCRs are an effective way of communicating nowadays. Much research is already done on the effect of OCRs on sales (e.g. Babic, et al., 2015; Floyd, et al., 2014). But not that much research takes a specific look to the tendency to engage in writing OCRs, especially taking into account the valence of their experience and the cultural background of the writer of the OCR. In this research we will get a better understanding of valence of a customer experience and the tendency to write an OCR.

1.2 Tendency to write an OCR

In this research we will get a better understanding about the motivations behind a consumer’s decision to engage in EWOM, and especially the tendency to write an OCR. The tendency to write a review is a motivational driver to write an OCR for different reasons, which will lead to different types of OCRs. Motives are defined as the ‘general drivers that direct a

(6)

consumers write reviews or read other consumers’ articulations in OCRs (Hennig-Thurau, Walsh & Walsh, 2003).

Given the conceptual closeness of traditional WOM and EWOM, the motives identified in the literature relevant for WOM can also be expected to be relevant for EWOM (Hennig‐Thurau, Gwinner, Walsh, & Gremler, 2004). Many research suggest that WOM mainly arise when consumers’ expectations are disconfirmed (e.g. Anderson, 1998), and other researchers found that the motives for engaging in positive WOM differ from motives which drives the negative WOM communication (Sundaram, Mitra & Webster, 1998). A few studies really examine in depth the underlying motivations of engaging in WOM, and even less studies examine explicitly the specific tendency of engaging in EWOM communication.

Most research investigates the drivers of motivation as brand-related factors such as

satisfaction, loyalty, quality, trust, perceived value, etc (de Matos & Rossi, 2008). This view, that brand experiences drive WOM is logical and managerially important. In literature this is also called situational- or product-involvement, which occurs when evaluating the product itself or having a short-term involvement in a product of low personal relevance (Bloch & Bruce, 1984; Wohlfeil & Whelan, 2006). But, this view does not account for drivers outside brand experiences. Some contextual factors also have been recognized in some literature, including: self-enhancement, concern for others, dissonance reduction (Engel et al. 1993; Hennig-Thurau et al. 2004); altruism, vengeance, and anxiety reduction (Sundaram et al. 1998). In this research we will get a closer look, building on het social psychology theory, to the more personal drivers to engage in WOM and write an OCR.

However, the motives to engage in EWOM are also effected by emotions of people (Berger and Milkman, 2012), and entails for instance achievement expressing, self confirmation and achieve a therapeutic feeling (Cheung et al. 2010). This is called enduring involvement, which relates to a more general attitude towards a product group or long-lasting involvement that arises out of a sense of high personal relevance (Bloch & Bruce, 1984; Wohlfeil & Whelan, 2006). One of the first extended research about emotional motives in WOM communication is done by Dichter (1966), they identified four main motivational categories of positive worth-of-mouth: product-involvement, self-involvement (self enhancement), other-involvement (concern for others), and message-involvement. These four major motivations for WOM were all about involvement. The typology of Dichter (1966) was modified after a few years by Engel, Blackwell, and Miniard (1993); they renamed the categories and added a new motive: dissonance reduction. After that, the most comprehensive study on motives in WOM

(7)

Based on the studies mentioned above, we examined variables which will be relevant to people’s tendency to write an OCR. There are several variables which can be identified as drivers from positive experiences, like; helping the company, altruism and self-enhancement. While there are some other variables which could be drivers from negative customer

experiences, like; anxiety reduction, warning other customers and vengeance. We use three broad variables to explain people’s tendency to write an OCR, and these three variables can be relevant for both positive and negative customer experiences. These variables which will be used in this research are: concern for others, self-enhancement and feedback for the company. In the next chapter these three variables will be explained more in depth. 1.3 Customer experience

Sunduram et al (1998) found in his research that motivation and experiences are closely related in the process of engaging in WOM. The underlying motives of customers to engage in WOM can be different, depending on the nature of their customer experience.

The whole idea of writing an online review will start with several customer experiences. The general concept ‘experience’ includes knowledge of or observation of objects or events through involvement in or exposure to them (Jeong & Jang, 2011). Zomerdijk and Vos (2010) conceptualize the experience as when a customer has any sensation of, or acquires knowledge from some level of interaction with elements created by the product or service provider. In this research we define these customer experiences as customers’ cognitive, affective, social, emotional & physical response to the total experience (Verhoef et al, 2009). The total

customer experience is from search, purchase and consumption, till the after-sales phase; having customers to sense, feel, think, act and relate to the company and his brands (Verhoef et al, 2009).

It will be interesting to investigate which tendency people have for writing an OCR, and if this tendency depends on the valence of the customer experience; positive vs. negative. Despite the considerable volume of research on eWOM, it will be important to knowledge that the tendency to engage in eWOM communications remains a really under-researched area in literature.

1.4 Valence and discrepancy of the customer experience

When people will purchase a product or make use of a certain service, they have already some expectations about what the experience will be for them (Woodruff & Gardial, 1996). It is possible that the actual experience not meet the prior expectation of the customer, so a difference or discrepancy will occur between the expectations and the real experience. The gap between prior expectations and the real experience will lead to the

Expectancy-Disconfirmation Paradigm. This Paradigm referrers to a disconfirmation of expectations (Rust & Oliver, 1997) and implies that the pre-purchase expectations become a standard to which post-purchase performance is contrasted (Reisinger & Waryszak, 1996; Tribe &Snaith, 1998). Within papers about customer experience and service quality, disconfirmation has been conceptualized in terms of meeting product or service attribute expectations (Oliver &

(8)

are presented; (1) positive disconfirmation, where performance exceeds expectations, (2) confirmation, where performance equals expectations, and (3) negative disconfirmation, where performance falls below expectations. These customers’ positive and negative experiences will be designated by their subjective judgment of the current experience and previous experiences (Jeong & Jang, 2011).This suggests that the prior expectations can lead to a certain positive or negative customer experience, which in turn influences people post-purchase behaviour like WOM or EWOM. East et al. (2007) suggest that the motivations behind a consumer’s decision to engage in positive vs. negative WOM communication are likely to differ. Therefore, we specifically separate negative and positive eWOM

communication into two different concepts. Customer experiences can led to the feeling of satisfaction or dissatisfaction related to the firm or brand, this depends on if the experience was valued as positive or negative. The valence of the customer experienceswill be relevant because prior experience influence future experiences of customers (Verhoef et al., 2009). Anderson (1998) found in his study that customers with extreme levels of satisfaction or dissatisfaction are more likely to share experiences than customers with a neutral opinion. So, there will be a certain level of discrepancy in people’s experience with a product or service. Also, a strong emotional imbalance, known as schema discrepancy, is an important antecedent for engaging in word-of-mouth, which considered triggering consumers to write an OCR (Ullah, Amblee, Kim & Lee, 2006).

1.5 OCR and different cultural backgrounds

The internet creates a medium which can reach audiences directly and generate buzzes with tremendous efficiency and flexibility, regardless of geographic boundaries. So, people all around the world are exposed to online customer reviews. Although many studies explained the impact of culture on marketing, few have examined the effect of cultural values on consumers’ WOM (Money, Gilly, and Graham 1998), and even less on OCRs. So, it will be interesting to research a cross-cultural investigation on OCRs, to explore if people of different backgrounds have different motivations regarding the tendency to write OCRs. This will be interesting because people all around the world are exposed to online customer reviews, but everyone can have a different view on these reviews and can give feedback of interpret in different ways. Culture is a broad topic in literature and can be defined as “shared motives, values, beliefs, identities, and interpretations or meanings of significant events that result from common experiences of members of collectives and are transmitted across age generation” (House, et al., 2004). Hofstede (2001) stated that culture is ‘the collective programming of the mind, which distinguishes the members of one group from another. We can suggest out of theses definitions that people acquire culture from their society or group of affiliation. So, people with different cultural backgrounds have their own way to look at things in life, which learned, shared, and defines the boundaries of different groups of people (Hall, Beyond Culture, 1976).

(9)

expected to choose different groups, messages, and methods to effect transmission (Chow, Deng, and Ho, 2000; Kale, 1991). So, in line with that reasoning we can imagine that different cultural backgrounds look in different ways at OCRs, and so strongly influence the tendency to write OCRs in different ways. In this research we want to investigate if there are difference between groups with different cultural background within the OCR field, especially in the relation of valence and discrepancy on the tendency to write an OCR.

1.6 Managerial and academic relevance

“Knowledge is power” can be applied in this study. There are a few studies which actually undermine in depth the tendency of engaging in WOM, and even less examine explicitly the more personal tendency to write an OCR. It will be relevant for managers to know why their customers engage in writing OCRs after their experiences. What exactly their tendency will be to write their good or bad experience on the internet. Also it is interesting to look at the schema of discrepancy in customer experiences and those effects on the tendency to write en OCR, this is namely considered as one of the main triggers to customers to engage in OCRs (Ullah, Amblee, Kim & Lee, 2016).

Limited prior research takes a specific look on the tendency to engage in writing OCRs, especially taking into account the valence of their experience and the cultural background of the writer. In this research, the researcher will attempt to fill this gap and expand the

knowledge in the field op tendency to write OCRs.

For marketing managers it will be interesting which type of discrepancy lead to which tendency to write an OCR. So, the get more knowledge about the motivation behind the customer review, this for self-enhancement, concern for others or feedback to the company. This knowledge is important because the posted OCRs are often basis for other customers to base their purchase decisions on (Zhu & Zhang, 2010), and this has an effect on the future sales of the company (Forman et al., 2008). And the thinking that negative WOM always negatively affects other customers and thus negatively effects the profit for the firm can be seen in a different way. It is also possible that a customer writes a negative OCR with the tendency to give feedback to the company, do the company should elaborate on this and take advantage of it.

In order to predict the review writing behaviour of customers, marketing managers should look at the tendency of these customers to write and post an OCR. This could give valuable insights into the different intentions to write an OCR and what experience it comes from.

1.7 Problem statement & research questions

EWOM is a very popular topic at the moment, and so are OCRs. As a result, many

(10)

discrepancy of the customer experiences. Secondly, it will be interesting to investigate the effect of cultural differences on the tendency to write these different OCRs. The earlier stated issues lead in this paper to the following research question:

“What is the impact of the valence and discrepancy of a customer experience on the tendency to write online customer reviews stemming from different motives, taking into account the different cultural backgrounds of the customers. “

This extended research question above can be decomposed in several sub-questions, these questions will be further explained in the next chapter.

RQ1: What is the impact of the valence of a customer experience on the different tendencies to write an OCR (concern for others, feedback for the company and self-enhancement)?

RQ2 : What is the impact of discrepancy of a customer experience on the different tendencies to write an OCR (concern for others, feedback for the company and self-enhancement)?

RQ3: What is the impact of the interaction of the valence en discrepancy of a customer experience on the different tendencies to write an OCR (concern for others, feedback for the company and self-enhancement)?

RQ4: What is the impact of cultural background on the different tendencies to write an OCR (concern for others, feedback for the company and self-enhancement)?

RQ5: RQ1, RQ2, RQ3, with taking into account the cultural background of the potential writer of the OCR.

1.8 Structure of the thesis

(11)

2. Theoretical framework

The previous chapter provided some general information about the concepts used in this research. Also, the managerial and theoretical relevance was mentioned why it will be interesting to do research in this specific field. The next chapter will go in more detail to the used concepts of this thesis in the theoretical framework of the paper. There will be explained what variables are used and which hypotheses are formed based on findings in previous literature.

First, the dependent variable (tendency to write an OCR) will be explained and discussed. Second, there will be more information given about the independent variables of this research (valence and discrepancy of a customer experience). Thereafter, the moderating variable will be explained (cultural background), and at the end of the chapter there will be a conceptual model including the hypotheses of this research.

2.1 Dependent variable: Tendency to write an OCR

The different customer experiences will lead to a specific tendency to share this experience via writing an OCR. In this research we will get a better understanding about the motivations behind a consumer’s decision to engage in EWOM, and especially the tendency to write an OCR. Research by East et al (2007), shed some light on the issue that the motivation for individuals to engage in positive word-of-mouth (PWOM) is likely to be different from that of negative word-of-mouth (NWOM) communications. Despite the considerable volume of research on EWOM, it will be important to knowledge that the tendency to engage in EWOM communications remains a really under-researched area in literature.

Based on the studies mentioned earlier in the introduction, we examined variables which will be relevant to people’s tendency to write an OCR. There are several variables which can be identified as drivers form positive experiences, like; helping the company, altruism and self-enhancement. While there are some other variables which could be drivers form negative customer experiences, like; anxiety reduction, warning other customers and vengeance. We use three broad variables to explain people’s tendency to write an OCR, and these three variables can be relevant for both positive and negative customer experiences. These variables which will be used in this research are: Concern for others, self-enhancement and feedback for the company. We use these three tendencies because they were used in many papers before about motivations to write an OCR. And also important, these three tendencies can be relevant for both positive and negative experiences.

1. Concern for others

(12)

other people. So, concern for others implies an element of altruism, which implies a voluntary act benefiting others without expecting something in return. When we look at this in the light of positive vs. negative experiences we can make a distinction between those. We can

differentiate it according to a positive customer experience in ‘helping other customers’, likely to show a concern by helping others to make a right decision, for instance telling that the restaurant was really worth to visit. And secondly differentiate it according to a negative, unfavourable customer experience in ‘warning other customers’, likely to show a concern to warn other people to prevent them from experiencing the same problems (Fu, Ju, Hsu, 2015). So, a positive or negative experience can both lead to a tendency ‘concern for others’ to write an OCR.

2. Feedback for the company

The underlying meaning of this concept is the same as ‘concern for others’; altruism or a desire to help others (Jeong & Jang, 2011). The difference between feedback for the company and concern for others will be the object. When we look at this concept in the light of positive versus negative experiences we can make a distinction between those. Positive customer experiences will lead logically to positive feedback for the company, when they firm equals customer expectations or the performance exceeds their expectations they will place an OCR like; ‘It was really good, keep going!’ A very positive experience creates a consumer’s true wish to help and promote the company (Riivits-Arkonsuo and Leppiman 2013). Negative customer experiences can both lead to positive and negative feedback for the company. Positive feedback could be some critic on the product or service and some recommendations to improve it, but negative feedback to the company could be; ‘You could better close the firm if you have this way of service!’

3. Self-enhancement

In the article of Stern (2000) is stated that self-enhancement includes egoistic value orientations. Egoism is defined as the ultimate goal is to enhance one’s own welfare or interest (Cheung & Lee, 2012). People with self-enhancement values desire positive

(13)

4. Relation between independent variables and dependent variable.

We argue that these tendencies to engage in writing an OCR are not equally distributed between the negative and positive customer experiences. So, we want to investigate what the differences in tendency will be between negative and positive customer experiences.

Secondly, we investigate if there are differences or similarities between the degree of discrepancy of the positive and negative customer experiences.

2.2 Independent variable: Valence of the customer experience.

Every customer is different, has different expectations and so also different experiences with a specific company, service or product. These customer experiences will lead to certain

feelings of satisfaction of dissatisfaction, depending on if the experience was valued as positive or negative (East, et al., 2007). And these specific customer experiences will lead to the tendency to share their experience and write an OCR. As we already mentioned the motivations behind consumer’s decisions to engage in WOM communication are different (East, et al., 2007), and this sharing of experiences can either be positive or negative (Buttle, 1997).

There is already a lot of research done in the field of when people engage in WOM or EWOM (e.g. East, et al., 2007; Floyd, et al., 2014). But if people with a negative experience engage more or less in WOM than satisfied people is supported for both possibilities in literature (Yi, 1991). In the literature is found that positive or negative experiences will lead to a tendency to write an OCR, but there is no clear difference yet in explaining which valence will give the highest tendency. Previous literature about the valence of the customer experience in relation to WOM found that there was no clear difference between the tendencies to engage in WOM. People with a positive experience, who are satisfied with the firm will engage in WOM favourable to the firm (Bitner, 1990; Reichheld & Sasser, 1990). Holmes and Lett (1977) found that satisfied customers engage more in WOM than dissatisfied customers. A greater tendency to engage in WOM is also found to occur as a result of greater satisfaction because of complaint handling (TARP, 1981). Also a great extent of literature research provide evidence for a reversed relation, that people with a negative experiences who are dissatisfied has a stronger tendency to engage in WOM (Richins, 1983; Westbrook, 1987). Like in an early customer oriented study, they showed that dissatisfied customers engage twice as much in WOM than satisfied customers (TARP, 1981). But the difference in valence of a customer experience can lead to different kind of tendencies to write an OCR, like concern for others, feedback to the company or self-enhancement.

(14)

2.3 Independent variable: Degree of discrepancy of the customer experience We also want to investigate the effect of the discrepancy of the customer experience in general and the discrepancy within positive and negative customer experiences. This will be interesting because an important antecedent for engaging in WOM will be a strong emotional imbalance known as schema discrepancy, which considered to trigger consumers to write an OCR (Ullah, Amblee, Kim & Lee, 2016). Anderson (1998) said that the more extreme the experience, the more people are intent to engage in WOM.

H2: High discrepancy of a customer experience is positive related to the tendency to write an OCR for CFO/FC/SE

And even more interesting to investigate will be interaction effect of the degree of

discrepancy and the valance of the customer experience. When we look at the discrepancy of the valence of a customer experience we can get 4 possible options: 1) A highly positive experience; 2) A moderate positive experience; 3) A highly negative experience; and 4) A moderate negative experience. In the research of Anderson (1998) they show a U-shaped relationship between customer satisfaction and the tendency to engage in WOM. This

suggests that extremely positive and extremely negative customers are most likely to write an OCR. In graph is shown that we expect that it will look like an inverted U-shaped relation.

H3: The negative valence which has a positive effect on the tendency CFO/FC/SE will be stronger with a high level of discrepancy

2.4 Moderator variable: Cultural background

People all around the world are exposed to EWOM, and more specific for this research to OCRs. Some studies researched the effect of cultural values on WOM (Money, Gilly & Graham, 1998), but less studied that effect on EWOM and OCRs specifically. In several studies it is mentioned that culture has a significant impact on customer behaviour, so it will be interesting to investigate culture in the light of OCRs. In this research we want to

investigate if people of different cultural backgrounds have a different tendency to engage in writing an OCR.

The Hofstede model is still one of the most popular models used to examine the effect of culture in marketing activities (e.g., Fam and Merrilees 1998; Liu, Furrer, and Sudharshan 2001). The Hofstede dimensions have also been identified as being particularly relevant to the cross-cultural investigation of service failure, recovery, and associated behaviours (Patterson et al., 2006).

2.4.1 Tendency to write an OCR

(15)

cultures people belong to in-groups which look after them (Hofstede, 2001). The more Western countries represent mostly the more individualistic culture (e.g. The Netherlands, Germany) and Eastern countries are more collectivistic (e.g. China, Russia) (Hofstede, 2001). Individualistic people consider themselves often as independent form others, give personal goal priority, emphasize rationality, and attitudes determine social behaviour. Collectivists consider the self as interdependent of others, give in- group goals priority, emphasize relatedness, and norms determine social behaviour (Triandis, 1995; Triandis & Gelfland, 1998). Both cultural dimensions are extremes from each other, and it will be interesting to look at these extremes in combination with the tendency to engage in OCRs.

Previous researchers suggests that collectivistic cultures show greater reliance on personal sources of information which encourage group information sharing, thus will lead to a greater tendency to engage in private WOM with family and friends (Liu and McClure, 2001; Babin et al., 2005; Patterson et al., 2006). But customers form individualistic cultures are seen as people who have higher levels of information giving (Fong and Burton, 2008), a greater willingness to provide recommendations and opinions (Babin et al., 2005), and to engage in negative word-of-mouth (Liu et al., 2001). Individualistic cultures express their opinions to define themselves, by making their thoughts tangible they can show other people their identity (Kim & Sherman, 2007). Individualistic people express their own personal opinion, even though this might lead to a direct conversation (Hofstede, 2001). In collectivistic cultures it will be the opposite, opinions are pre-determined by in-groups and the harmony should

always be maintained (Hofstede, 2001). So, collectivistic people can be less willing to express their opinions, this because it might lead to direct confrontation, which should be avoided (Kim & Sherman, 2007; Hofstede, 2001). Lai et al. (2013) suggests that individualistic seems more willing and likely to express their opinion in online reviews than collectivists.

H4a: Individualistic cultures have a positive effect on the tendency to write an OCR for CFO/FC/SE.

Another interesting cultural dimension related to the tendency to write OCRs is

masculine/feminine. More masculine cultures include values like the importance of showing off, performing, achieving something visible, making money and big is beautiful (Hofstede, 1983). These cultures are characterized by a stronger ego orientation, such that people define themselves and their reason for being according to their work and money or belongings. The more feminine cultures include values like not showing off, putting relationships with people before money, minding the quality of life and the preservation of the environment, helping others (in particular the weak), and small is beautiful (Hofstede, 2001). These cultures are characterized by a stronger relationship orientation, the quality of life and people are more important. Because of the more materialistic and possession-oriented nature of masculine cultures, researchers found evidence for higher levels of information sharing (Dwyer, Mesak & Hsu, 2005; Lam, Lee & Mizerski, 2009; Liu, Furrer & Sudharshan, 2001).

(16)

2.4.2 Valence and the tendency to write an OCR

Stated is that individualistic people engage more in writing an OCR when face a negative experiences, compared to collectivistic cultures. This because collectivistic cultures often don’t repressed outwardly negative emotions (Markus & Kitayama, 1991). Research of Cheung, Anitsal and Anitsal (2007) a comparison of China and the U.S. was conducted on word-of-mouth tendencies. They found that individualists tend to be strongly motivated to initiate in negative WOM. As for the collectivistic people, they were found to be equally motivated to initiate both positive and negative WOM; when talking about sharing positive experiences, they never hesitate (Cheung, Anitsal & Anitsal, 2007). In contrast, it is important for collectivistic people not to give the in-group members the same bad experience they had before (Liu & McClure, 2001). Individualistic customers are much more likely to voice their complaints than collectivistic customers (Liu & McClure, 2001). This because individualistic customers are concerned with their personal needs, goals, and interests (Chelminski & Coulter, 2007). They feel responsible to make complaints when they are dissatisfies in order to improve their quality of life (Ngai et al., 2007).

H5a: The positive effect of a negative customer experience on the tendency CFO/FC/SE will be stronger for more individualistic cultures.

There is a lot of research about the cultural dimension individualism/collectivism and engaging in positive or negative EWOM, but still less literature about this relation with masculine/feminine. When we get a look at the cultural dimension masculine/feminine we see that the more masculine cultures value showing off and have a stronger ego orientation (Hofstede, 1983). These characteristics of masculine say that they are more self-confident and assertive (Lam, Lee & Mizerski, 2009). In the literature about complaining behaviour

Donoghue and de Klerk (2006) found that people who complain often are more assertive and self-confident. Therefore we propose the following hypothesis:

H5b: The positive effect of a negative customer experience on the tendency CFO/FC/SE will be stronger for more masculine people.

2.4.3 Degree of discrepancy and tendency to write an OCR

When the discrepancy of a customer experience get higher, also the tendency to write an OCR goes up. But this effect could be weakened or strengthened by the dimensions

collectivism/individualism and masculine/feminine. As stated before, individualistic seems more willing and likely to express their opinion in online reviews than collectivists (Lai, et al., 20013). Collectivistic people are less likely to express their feelings and opinions because they would avoid direct confrontation (Kim & Shermand, 2007). But, collectivistic people are more concerned for the society and others, so they would share opinions with a high

(17)

H6a:The positive effect of high discrepancy of a customer experience on the tendency CFO/FC/SE will be stronger when people are more collectivistic.

For the dimensions masculine feminine we could say that when people are more masculine they are self-confident and assertive, and would give their opinion about a certain experience. It would not matter if the discrepancy of this experience would be high or low, so the positive effect of high discrepancy on the tendency to write will be lowered, so it becomes more equal to low discrepancy.

H6b: The positive effect of high discrepancy of a customer experience on the tendency CFO/FC/SE will be lower when people are more masculine.

2.5 Control variables

In accordance with prior literature two control variables will be included in the model. These variables are assertiveness and concern for face.

2.5.1 Assertiveness

It will be interesting to take into account the personal value assertiveness as a control variable in this research. This because it can be an important determinant within the cultural dimension masculinity/femininity. Lam, Lee and Mizerski (2009) mentioned in their research that

masculinity is a preference for valuing assertiveness, achievement, and material success. Assertiveness is the degree to which individuals are assertive, confrontational, and aggressive in their relationships with others (House et al., 2004). So, we need to know is the dimension masculine/feminine is the driver, or if assertiveness will be the driver within these dimension, so it is needed to control for that possibility.

2.5.2 Concern of face

The concept ‘concern for face’ is not so much what one claims for oneself, but rather the image one wishes to project concerning ones social position, status, and credibility (Chen, 2013). It is the positive way people want others to see them (Locher & Watts, 2005). In literature is found that face concern can influence customer satisfaction and dissatisfaction responses, such as complain intention or pleasure sharing (Chan &Wan, 2009). Hui and Triandis (1986) argue in their paper that collectivists are more concerned of face than

(18)

2.6 Conceptual model

According to previous theory and hypotheses which describe the effects of the independent and moderating variables on the dependent variables, the conceptual model is created (Figure 1). The three dependent variables are shown; tendency to write because of concern for others, feedback for the company and self-enhancement. The two independent variables are

presented; valence of the customer experience and the degree of discrepancy of the customer experience. There is also an interaction effect expected between the valence and discrepancy of the customer experience on the dependent variables. Furthermore, the cultural dimensions as moderator variables are shown; masculine/feminine and individualism/collectivism. The arrows represent the different hypotheses about the variables. These variables are tested as direct effect on the tendencies to write an OCR. The control variables are also included in this conceptual model by a dotted line, this because we control for the effect of these variables, but we do not include these relations in certain hypotheses. Below the conceptual model in table 1 are the hypotheses of this research summarised. Presented in the following chapter are the methodology and measures which will be used for this research.

(19)

Table 1: Hypotheses for the three tendencies to write an OCR: CFO, FC and SE Hypothesis 1 A negative valence of a customer experience is positive related to the

tendency to write an OCR for CFO/FC/SE.

Hypothesis 2 High discrepancy of a customer experience is positive related to the tendency to write an OCR for CFO/FC/SE

Hypothesis 3 The negative valence which has a positive effect on the tendency CFO/FC/SE will be stronger with a high level of discrepancy

Hypothesis 4a Individualistic cultures has a positive effect on the tendency to write an OCR for CFO/FC/SE.

Hypothesis 4b Masculine cultures has a positive effect on the tendency to write an OCR for CFO/FC/SE

Hypothesis 5a The positive effect of a negative customer experience on the tendency CFO/FC/SE will be stronger for more individualistic cultures.

Hypothesis 5b The positive effect of a negative customer experience on the tendency CFO/FC/SE will be stronger for more masculine people.

Hypothesis 6a The positive effect of high discrepancy of a customer experience on the tendency CFO/FC/SE will be stronger when people are more collectivistic. Hypothesis 6b The positive effect of high discrepancy of a customer experience on the

(20)

3. Research methodology

The methodology used is explained in this chapter, this by outlining the research that has been done and describing the data that resulted from the distributed surveys. This with the aim to test the hypotheses based on theory in the previous chapters. In the first section, the research design will be described and thereafter we will elaborate on the participants of the research. In the last section we will discuss the actual plan of analysis of this research, including the measurements of the concepts and planned process of the data analysis.

3.1 Research Design

In this research we will get information about the research questions by manipulation of the independent variables, so an experimental design will be used. This design is the main method used in research conducted to obtain evidence of a cause-and-effect relationship (Malhotra, 2010). The experiment can be described as a 2 (valence of experience: positive/negative) x 2 (degree of discrepancy: low/high) x 3 (tendency to write an OCR: concern for others/feedback for company/self-enhancement) between-subjects factorial design (Malhotra, 2010). So, there are two independent variables: valence of the customer experience and degree of discrepancy of this experience, and one dependent variable: tendency to write an OCR. A between-subject design is chosen to avoid carry-over effects (Malhotra, 2010); every respondent is only

assigned to one of the four conditions.

(21)

broad range of cultural different between those nations, still in this research we will get a look at the remaining differences between those countries. Furthermore, we include China in this research about customer reviews. This because many research stated that Western cultures are really different from Asian cultures (e.g. Hofstede, 2001; Schwartz, 1994), but when we get a look at the masculine/feminine dimension, China is actually similar to the western country Germany. Relating to the dimensions of Hofstede we could say that The Netherlands and Germany are reasonable similar on the cultural dimension individualism/collectivism, but China and Germany are equal on the dimension masculine/feminine (Figure 2; Table 2). We try to find an equal distribution of participant of the three countries to investigate; Germany, the Netherlands and China. These three groups of participants will be equally distributed among the four different scenarios in the survey.

Figure 2: Culture dimensions of Hofstede

Table 2: Culture dimensions of Hofstede

Country Individualism/Collectivism Masculine/Feminine

The Netherlands Individualism Feminine

Germany Individualism Masculine

(22)

3.1.1 Participants

A total of 295 users participated in this study, whereof 255 responses are used for the

analysis. This difference between total and used responses of the survey is due to the fact that some participants didn’t completed the survey in a sufficient way. In order to get a reliable examination of the independent variables on the dependent variable, we need approximately 20 respondents of each of the three countries for the four different conditions. This will add up to a total of 240 respondents needed for this research.

Table 3 summarises the demographical statistics and characteristics of the 255 participants. 147 respondents are female (i.e. 61.5 %) and the largest group of participants belongs to the age group 22-30 years (75%). Most of the respondents of this survey have their Bachelor degree (45.8%). The nationality of the participants in this survey is distributed like, 29.8% German, 32.9% Chinese and 37.3% Dutch respondents. This distribution is fine to investigate these three countries among the four different scenarios in this research.

Table 3: Demographical statistics (N=255)

Frequency Percentage Gender Male 92 38.5% Female 147 61.5% Total (N) 239 100% Age < 22 years 30 12.5% 22-30 years 180 75% 31-45 years 24 10% 46 > years 6 2.5% Total (N) 240 100% Nationality German 76 29.8% Chinese 84 32.9% Dutch 95 37.3% Total (N) 255 100%

Education level High school diploma 50 20.8%

Trade/Tech/Voc training (MBO) 2 0.8%

Associate degree (HBO) 20 8.3%

Bachelor degree 110 45.8% Master degree 47 19.6% Doctorate degree - 0.0% Other 11 4.6% Total (N) 240 100% 3.1.2 Survey design

(23)

groups, and will be distributed by convenience sampling (Malhotra, 2010). This means that the target population for the survey will be every person who has access to the Internet, especially access to e-mail and social media. The choice for working with a survey via internet is because, the internet is the fastest method of obtaining data from a large number of respondents and has really low costs (Malhotra, 2010). And even more important, the research area concerns OCRs on the internet, so it is necessary to reach participants who actually use the Internet and are known with writing a review on the Internet. To ensure that enough people should fill in the survey, convenience sampling is used because the easy occurrence of the snowball effect (Malhotra & Birks, 2007). The receivers of the message to fill in the questionnaires are asked to forward the survey and spread the word to fill in this survey, additional respondents are collected by previous respondents. Also, printed surveys are used and handed out to several respondents to fill in the questionnaire. The reason for this were time constraints and difficulty to find enough German and Chinese respondents, who could sufficiently complete the survey. These respondents are randomly chosen, only based on their nationality. This process has taken place in the Zernike Duisenberg building, due to the fact that there are a lot international students studying, so these people could be easily reached and asked to fill in the survey. The questions are asked in a predetermined order, this to enhance the reliability, and the questions are in English since the respondents will be from different countries or regions. There is no incentive offered to the respondents, so the participation in this research is completely voluntary.

3.1.3 Scenario analysis

In order to effective study the effects of the degree of discrepancy on the tendency to write an OCR with taking into account the valence of the experience, both the degree of discrepancy and the valence of the customer experience are manipulated. To measure these tendencies it is important that there will be certain contact with the customer right after the experience, in this way the customers still have a clear thought about the specific experience they had (Huss, 1988). Consumers will express their feelings about certain experiences online by writing a review, and the best way to get the right reaction and feelings about the experience is to directly confront them about it. With that method it is possible to collect very specific

individual data, but it would be very difficult to find a large sample and impossible in the time span of this study. To eliminate these problems we make use of a scenario analysis in this study, using a between-subjects design.

Huss (1988) described a scenario as a ‘narrative description of a consistent set of factors which define in a probabilistic sense alternative sets of future business conditions’. Scenario analysis is part of a vignette study. These studies have an experiential design where

respondents are exposed to similar, but not identical scenarios (Huss, 1988). These vignettes are short descriptions of situations, persons of attributes, shown to respondents to elicit their judgements about a scenario (Alexander & Becker, 1987). The purpose of these kinds of studies is to examine the intentions and actions of individuals who read the specific scenarios (Wallander, 2009). These vignettes are often used to provide realism, improve construct validity, reliability and time constraints when analysing real experiences (Wason, Polonsky & Hyman, 2002).

(24)

experience. These different scenarios contain the valence of the experience and the

discrepancy of that experience; this will result in four different scenarios, which are given in table 4. Each of the scenarios differs in either the degree of discrepancy or valence. The participants will be asked to imagine if they had that specific customer experience. These four different scenarios have all the same design and are all written about the same product and products characteristics.

The four scenarios were equally distributed among the three nationalities (German, Chinese, Dutch), and one of these scenarios was assigned to one individual participant. By using options in Qualtrics, the four scenarios have almost equal sample sizes across the different cultures. The descriptive per scenario are shown in the tables 5, 6, 7 and 8.

Table 5: Descriptive statistics of high discrepancy with a negative customer experience (--) Educational background***

Total M/F* Av. Age** HS MBO HBO Bd Md Dd Other

Dutch 19 8/11 22 2 - 4 10 3 0 -

German 15 9/6 23 4 - 1 6 3 0 1

Chinese 21 12/9 24 5 - 1 12 2 0 1

Total 55 29/26 23 11 - 6 28 8 0 2

*Percentage Male (M) and Female (F) **Average age of the respondents

***Different educational background. HS: High school diploma; MBO: Trade/technical/vocational training; HBO: Associate degree; Bd: Bachelor degree; Md: Master degree; Dd: Doctorate degree; Other.

Table 6: Descriptive statistics of low discrepancy with a negative customer experience (-) Educational background***

Dutch 21 4/17 22 1 1 4 9 6 0 -

German 17 6/11 22 7 - 1 7 - 0 2

Chinese 25 8/17 24 6 - 2 8 7 0 2

Total 63 18/45 23 14 1 7 24 13 0 4

Table 4: Scenarios containing valence and discrepancy of the customer experience

-

Low discrepancy with a negative customer experience (It’s a little worse than I expected)

+

Low discrepancy with a positive customer experience (This exceeds my

expectations a little) --

High discrepancy with a negative customer experience (This is much worse than I expected)

++

High discrepancy with a positive customer experience (This exceeds my

(25)

Table 7: Descriptive statistics of low discrepancy with a positive customer experience (+) Educational background***

Dutch 23 5/18 23 5 - - 13 5 0 -

German 22 11/11 22 7 - 1 9 4 0 1

Chinese 15 8/7 24 2 - 2 6 5 0 1

Total 60 24/36 23 14 - 3 28 14 0 2

Table 8: Descriptive statistics of high discrepancy with a positive customer experience (++) Educational background***

Dutch 30 10/20 23 4 1 3 17 5 0 -

German 15 6/9 22.33 5 - 1 6 2 0 1

Chinese 16 5/11 34 2 - - 7 5 0 2

Total 61 21/40 11 1 4 30 12 0 3

3.1.4 Operationalization

This section will elaborate on the questions that are included in the questionnaire and which are used for subsequent analyses. The questionnaire that has been designed for the current research contains 49 questions.

It first starts with a short introduction of the survey, followed by the first question about the

nationality of the participant. This is the filter question of the survey. This question will

distribute the Chinese, German and Dutch participants equally over the given scenarios, and at the same time determines their cultural background.

(26)

For the given scenarios in the survey we made a choice between a hedonic or utilitarian customer experiences. For hedonic products or services, a customer review will be attributed to non-product related reasons such as personal taste of the writer. With utilitarian products of services, the consumption concerns maximizing utility, which is characterized by tangible and more objective criteria (Sen & Lerman, 2007). We choose for a utilitarian product, and this more specific will be a mobile phone in this research. This because utilitarian is more fact based and objective, compared to hedonic which experience can be different for every individual and is rather subjective. Specifically, consumer judgment with respect to these products tends to be cognitively driven, instrumental and goal-oriented, and accomplish a functional or practical task (Strahilevitz & Myers, 1998). We want to ensure that the give scenario will give the same experiential feeling among the respondents. And we choose for using a mobile phone in this research because we will expect that all of the respondents of this survey have a mobile phone and know what they value about it. This leads to the four

different scenarios explained in table 4 before, and are operationalized in table 9.

This scenario is followed by questions about the dependent variable; tendency to write an

OCR. The three tendencies which are used in this study are conducted from several other

studies which examined traditional and/or electronic Word-of-Mouth (Cheung & Lee, 2012; Hennig-Thurau et al., 2004; Yap, Soetarto & Sweeney, 2013). This to ensure that the items

Table 9: Operationalization of scenarios

Valence/Discrepancy High Low

Negative Scenario 1

--

You bought a mobile phone two weeks ago, and it is not working at all! It is far below your expectations. It crashes several times a day when you open just an application, the pictures made with the phone are of low quality, and the battery lifetime is really bad! Worst buy ever!

Scenario 2 -

You bought a mobile phone two weeks ago. It doesn’t meet your expectations and you are not that satisfied with it. It works well, but the battery lifetime is not what you expected and also the camera is not that good with taking pictures when facing less light.

Positive Scenario 4

++

You bought a mobile phone two weeks ago, and you are so happy with it! It exceeds your expectations! It works great, the battery lifetime is more than 2 days, and the camera makes beautiful pictures. Best buy ever!

Scenario 3 +

(27)

used to measure the tendency to write an OCR is valid and reliable, because these items are tested multiple times before in other relevant papers. The tendencies picked out of all this existing literature we will use for this research are: ‘concern for others’, ‘self-enhancement’ and ‘feedback to the company’. In table 10 is shown which items are used for the three tendencies and which literature is used for these items. The items are measured by using a seven-point Likert scale, with a range from “strongly disagree” to “strongly agree”. The whole list of the questions about the tendency can also be found in the survey in Appendix 1. After the questions about the dependent variable, a manipulation check will be done. This with the question; “How positive/negative do you perceive the given customer experience as portrayed before?”. With this check we will ensure that the respondents really get and understood the given customer experience.

Furthermore, the moderator variable is tested by certain questions about the

individualism/collectivism dimension, and the masculine/feminine dimension of Hofstede

(2001). Firstly, the respondent will face eight questions about the individualism/collectivism dimension, thereafter eight questions about the masculine/feminine dimension. These

questions are based in different scales and measurements of the articles of Yoo, Donthu & Lenartowicz, 2011, House et al., 2004, De Mooij, 2004, Hofstede, 2001 and Srite & Karahanna, 2006.

The last section of questions is concerned about demographics and other control variables. The first control questions are about the assertiveness of the respondents. Assertiveness is “the degree to which individuals are assertive, confrontational, and aggressive in their relationships with others” (House et al., 2004). The questions after that are about the topic concern for face (Cocroft & Ting-Toomey, 1994; White, Tynan, Galinsky & Thompson, 2004).

We think we need to take into account these control variables because the different behaviour of Chinese, German and Dutch people.

Thereafter we use the question; “In general, how often would you post an online customer review?” (Chiu et al., 2006) (Likert scale, ranging from 1 (never) to 7 (very often)).

This question will be included to understand if people will write OCRs in general, and if this influences their response of the survey.

Also two control questions are about the background of the participant;

“In which country were you born?” and “Where have you lived the most time of your life?”. These questions will be interesting if someone is from certain nationality, but is born of lived most of his/her life in another county. It is possible that this will affect and changed their behaviour and way of thinking, for instance the scores on the cultural dimensions

individualism/collectivism and masculine/feminine.

(28)

Table 10: Overview of items included in the questionnaire that are used for subsequent analyses

Variable Definition Items Scale Alpha

Tendency to write an OCR (depende nt variable)

Motives are defined as the ‘general drivers that direct a consumer’s behaviour attaining his or her needs (Assael, 1998). These motives significantly determine specific consumer behaviour and are therefore really useful in explaining why consumers write reviews; their tendency to write an OCR. (Hennig-Thurau, Walsh & Walsh, 2003).

Q4.1 - It is likely that I will write an online review about this specific experience. (Henning-Thurau et al., 2004) Likert scale, ranging from 1 (disagree strongly) to 7 (agree strongly) (Likert, 1932) Concern for others (depende nt variable)

An element of altruism, which implies a voluntary act benefiting others without expecting something in return (Henning-Thurau et al., 2004).

Q4.2 - It is likely that I will write a review to help other customers with this experience

Q4.3 - It feels good to help others with writing a review about this experience Q4.4 - It is likely that I will write a review about this experience to give others the opportunity to buy the right (Hennig-Thurau et al., 2004; Cheung & Lee, 2012; Yap, Soetarto & Sweeney, 2013) Likert scale, ranging from 1 (disagree strongly) to 7 (agree strongly) (Likert, 1932) α = .870 N=3 α = .892 N=2 Feedback for company (depende nt variable)

The underlying meaning of this concept is the same as ‘concern for others’; altruism or a desire to help others (Jeong & Jang, 2011). The difference will be the object; the company.

Q4.5 - I want to help the company to be successful with sharing this experience

Q4.6 - In my opinion, good companies should be supported with writing a review about this experience Q4.7 - I feel that companies deserve free publicity by writing a review about this experience

Q4.8 - I think it is legitimate to cause this specific (in)convenience by writing a review to a seller (Hennig-Thurau et al., 2004; Yap, Soetarto & Sweeney, 2013)

Likert scale, ranging from 1 (disagree strongly) to 7 (agree strongly) (Likert, 1932) α = .777 N=4 Self-enhance ment (depende nt variable)

The ultimate goal is to enhance one’s own welfare or interest (Cheung & Lee, 2012).

Q4.9 - My contribution in writing an review about this experience shows others that I am a knowledgeable customer

Q4.10 - It is likely that I want to express my feelings about this specific experience by writing an online review

Q4.11 - I feel good when I can tell others about this specific buying experience

(Hennig-Thurau et al., 2004; Yap, Soetarto & Sweeney, 2013)

(29)

Cultural backgrou nd (moderat or) – individua lism/colle ctivism

Cultural background: The collective programming of the mind, which distinguishes the members of one group from another (Hofstede, 2001). Individualism highlights cultures characterized by personal achievements, while collectivistic

cultures by group achievements and group loyalty (Hofstede, 2001).

Individualism/collectivism Q14.1 - Group success is more important than individual success. Q14.3 - Individuals should consider their goals after considering the welfare of the group.

Q14.5 - I feel good when I cooperate or spend time with others

Q14.7 - Being accepted as a member of a group is more important than being independent

Q14.8 - Being loyal to a group is more important than individual gain

Q14.2 - Individual rewards are more important to me than group welfare (reversed question)

Q14.4 - It is important to me that I do my job better than others (reversed question)

Q14.6 - I like to do just my own things (reversed question)

(Yoo, Donthu & Lenartowicz, 2011, House et al., 2004, De Mooij, 2004, Hofstede, 2001, Srite & Karahanna, 2006) Likert scale, ranging from 1 (disagree strongly) to 7 (agree strongly) (Likert, 1932) α = .747 N=5 Cultural backgrou nd (moderat or) – masculine /feminine

Cultural background: The collective programming of the mind, which distinguishes the members of one group from another (Hofstede, 2001). Masculinity represents cultures characterized by challenges and social achievements, while feminine cultures by quality of life, environmental care, security and attention to others (Hofstede, 2001).

Masculine/feminine

Q17.1 - It is more important for men to have a professional career than it is for women.

Q17.2 - Men usually solve problems with logical analysis; women usually solve problems with intuition. Q17.3 - Solving problems requires an active forcible approach, which is typical of men.

Q17.5 - Meetings are usually run more effectively when they are chaired by a man.

Q17.4 - There are some jobs that a man can do better than a woman Q17.6 - Achievement and success are NOT that important to me (reversed question)

Q17.7 - Taking care of others in the society is important to me (reversed question)

(30)

Control questions - Assertive ness Q23.1 - Non-assertive – Assertive Q23.2 - Non-dominant – Dominant Q23.3 - Undemanding – Demanding Q23.4 - Nonaggressive – Aggressive Q23.5 - Non-confrontable – Confrontable Q23.6 - Non-determined – Determined Q23.7 - Tender – Tough (House et al., 2004) 7-point Bipolar Scale α = .823 N=6 Control questions – Concern of face

Q26.1 - I care about others attitude towards me

Q26.2 - I am concerned with my social status

Q26.3 - I hate being taken lightly Q26.4 - I will be very angry if others are impolite to me

Q26.5 - I care about praise and critism from others

Q26.6 - I will be very happy if I am treated with respect

Q26.7 - I am concerned with my self-image

Q26.8 - I will be very upset if I am criticized in public

(Cocoft & Ting-Toomey, 1994; White, Tynan, Galinsky & Thompson, 2004)

Likert scale, ranging from 1 (disagree strongly) to 7 (agree strongly) (Likert, 1932) α = .814 N=8 Control questions – Demogra phics

- In which country where you born?

- Where have you lived in the most time of your life? - What is your gender? - What is your age?

- What is the highest level of education you have

(31)

3.3 Dimensions of Hofstede

Before diving into testing of the hypotheses, we need to explore if the participants of the chosen nationalities in this research are actually that different as explained in the literature. With an ANOVA and a post hoc analysis we will analyse if these nationalities are significantly different from each other on the dimensions of Hofstede (2001); individualism/collectivism (I-C) and masculine/feminine (M-F). Also we take into account the control variables assertiveness (ASS) and concern for face (CF), in these analyses. In table 11 the descriptive information of the ANOVA of the variables is shown, and in table 12 we can see which nationalities are significant different from each other on with cultural dimensions.

First, on the dimensions I-C, we can conclude that only Dutch and Chinese people (p=.021) significantly differ from each other. This is in contradiction with the literature, where is stated that both the Netherlands and Germany are seen as more individualistic cultures, and China as a more collectivistic country.

Second, on the dimensions M-F we see that both German and Chinese (p=.002), and Dutch and Chinese (.000) people are significantly different from each other. In the literature in mentioned that the Netherlands is a more feminine country, and Germany and China would score higher on the masculine dimensions. These results suggest that China is more masculine than both the Netherlands and German, but German is not significantly different from the Netherlands (p=.994). The German participants of this research are more feminine than stated in the literature.

After the cultural dimensions of Hofstede (2001), we used some other cultural dimensions in this research as control variables. On the dimensions ASS, all three nationalities differ significantly from each other, where Dutch people are most assertive (M=4.81) and Chinese people has the lowest score on this dimension (M=3.88). On the dimensions CF only the difference between Dutch and Chinese participants was

significant in this research (P=0.53). The Chinese participants are more concerned of face (M=5.28), than the Dutch participants (M=4.96).

Concluding we can say that the group participants of this research score different on some cultural

dimensions than in literature is stated. This can be due to the small group of participants (N=255), around 80 people for each nationality could be too less to have significant results on all cultural dimensions and it is not a perfect representation of all people of that specific nationality. Also it can be due to that most participants are in the age range of 22-30 years, so younger people. Since the research papers about the cultural dimensions of Hofstede (2001) a lot has changed in the world, so it could be that people do not response in the same way if it was in the early years of 2000 and before.

Table 11: Descriptives of cultural dimensions Nationality

participants

Individualism/collectivism Feminine/masculine Assertiveness Concern of face

N M SD N M SD N M SD N M SD

German 75 4.92 0.96 75 2.99 1.49 65 4.39 0.86 70 5.09 .088

Chinese 80 5.06 1.10 80 3.78 1.57 77 3.88 1.07 77 5.28 1.02

Dutch 94 4.65 0.90 93 2.92 1.13 93 4.81 0.93 93 4.96 0.76

Table 12: Significance level of the multiple comparison of the nationalities

Individualism/collectivism Feminine/masculine Assertiveness Concern of face

German – Chinese .672 .002 .005 .432

German – Dutch .182 .944 .023 .590

The relation between the tendency to write an online customer review and the discrepancy of a positive or negative experience in a cultural perspective.